The invention relates to the field of magnetic transducers (heads) having inductive write heads and more particularly to the structure of and the process for making the pole pieces and coils for the inductive write head.
A typical prior art disk system 10 is illustrated in FIG. 1. In operation the magnetic transducer 20, usually called a xe2x80x9cheadxe2x80x9d is attached to an arm or actuator 13 and flies above the rotating disk 16. A voice coil motor 19 (VCM) pivots the actuator 13 to position the magnetic transducer 20 over selected circumferential tracks on the disk 16. The disk 16 is attached to spindle 18 that is rotated by a spindle motor (not shown). The disk 16 comprises a substrate on which a plurality of thin films are deposited. The thin films include ferromagnetic material that is used to record the magnetic transitions written by the magnetic transducer 20 in which information is encoded. A tape based storage system (not shown) uses a magnetic transducer in essentially the same way as a disk drive, with the moving tape being used in place of the rotating disk 16.
The magnetic transducer 20 is composed of elements that perform the task of writing magnetic transitions (the write head 23) and reading the magnetic transitions (the read head 12) as illustrated in FIG. 2. The electrical signals to and from the read and write heads 12, 23 travel along conductive paths (leads) (not shown) which are attached to or embedded in the actuator 13. Typically there are two leads each 14 for the read and write heads 12, 23.
FIG. 2 is a midline section of one type of prior art magnetic transducer 20A. The components of the read head 12 are the first shield (S1), two insulation layers 107, 109 which surround the sensor element 105 and the second shield 104 (P1/S2). This type of magnetic transducer 20A is called a xe2x80x9cmerged headxe2x80x9d because the P1/S2 layer 104 serves as a shield for the read head 12 and a pole piece for the write head 23A. The yoke also includes a second pole piece 103 (P2) which connects with P1/S2104 away from the air-bearing surface (ABS) at what is sometimes called the xe2x80x9cback gapxe2x80x9d (BG). The P2103 confronts the P1104 across the write gap layer 42 to form the write gap 43 at the ABS. The coil 37 in this particular prior art head is deposited on a layer of resist 106 which is used to define the zero throat height (ZTH) by forming a step on the gap layer 42.
FIG. 3 is a midline section of a second type of prior art magnetic transducer 20B. There are two significant differences between the magnetic transducers 20A and 20B in FIGS. 2 and 3. One difference is that the yoke in magnetic transducer 20B includes three pole pieces P1104, P2103A and P3103B. The P2103A is formed at the write gap 43 a separate element. The third pole piece 103B (P3) is stitched to P2103A and is connected to the P1104 at the back gap (BG) to complete the yoke. Typically write heads 23 only have one coil layer 37, but the particular write head 23B shown has two coil layers which be called coil137 and coil257. The turns of both coil1 and coil2 are routed between the write gap 43 and the back gap (BG) and then around behind the yoke. Coil1 and coil2 are connected electrically (typically behind the back gap) to form a single inductive coil. The P3103B arches over the resist mound 111 which surrounds the coil(s). In either of the prior art write heads 23A, 23B the angle at which the bottom surface of P2 moves away from the ZTH point is typically far less than 90 degrees which results in efficiency losses through flux leakage.
In either of the prior art heads 20A, 20B of FIGS. 2 and 3 additional coil layers can formed on top of the previous coils prior to forming the enclosing pole piece. Thus, three or more coil layers can be made within these basic designs. Adding additional coil layers, however, will not change the fundamental limitations on the yoke lengths in these heads.
As the required recording densities increase the width of the written track must decrease. The needed write heads must have high magnetic efficiency and low inductance. These requirements make it necessary to place the inductive components ever closer to the pole tips than is possible using the prior art.
In U.S. Pat. No. 6,194,323 to Downey and Yen, a process for making semiconductors is disclosed that uses a so-called hard mask. The hard mask is selected to be more resistant to the metal etchant being used, which in turn allows a thinner photoresist to be used with a resulting increase in resolution. The hard mask is deposited on the metal layer and a thin photoresist is deposited on the hard mask and patterned in the convention manner. The hard mask is then etched to expose portions of the metal layer that can then be etched to achieve the desired pattern of metal. The materials useful for the hard mask are said to include titanium nitride, silicon nitride, tungsten, titanium, various glasses, tantalum oxide, aluminum oxide, titanium oxide, as well as, organic hard masks such as spin-on anti-reflection coatings.
One process for forming a pattern of material with submicron dimensions is called the damascene process. In this method the pattern is developed by etching away selected dielectric material to form features (vias, troughs, etc.) that are then overfilled by electroplating a metal such as copper. The overfill is removed by chemical-mechanical polishing (CMP) leaving the metal and dielectric material forming the pattern. Hard masks have been used to improve the precision of the damascene process. In U.S. Pat. No. 6,121,150 Avanzino and Wang suggest use of sputter-resistant materials for the hard mask. Specifically, they teach the use of high atomic mass metallic materials such Ta, W, Ti, TaN, WN and TiN for the hard mask in a damascene process for fabricating semiconductors.
Applicants disclose a method for producing a magnetic transducer with a inductive write head having a multilayer coil with a high aspect ratio and a short yoke. A damascene process is used for two coil layers and a conventional process for the third coil layer. The process of the invention allows a seed layer for the coil to be deposited on the side walls of the trenches for the first and second coil layers. In one embodiment the seed layer for the coil is preceded by an adhesion layer.
In a preferred embodiment of the invention the first coil layer is formed below the gap layer and behind a pedestal pole piece for P1. A layer of polymeric material is deposited to serve as the bed for the turns of the coil. A layer stack including a hard mask layer followed by a masked photoresist layer defines the coil shape and enables etching (RIE) with a high aspect ratio. Tantalum oxide is preferred as the hard mask material. A seed layer is then deposited over the resulting trenches. This process results in the seed layer being deposited on the sidewalls as well as the bottoms of the trenches. In an alternate embodiment an adhesion layer may be deposited prior to the deposition of the seed layer. Tantalum is the preferred adhesion layer and copper is the preferred seed layer for the coil. A wet electroplating process is preferred to deposit the conductive material for the coil on the wafer to overfill the trenches. The first coil layer is completed when the overfill is removed by planarization. The gap layer is then formed. The second layer of the coil is formed above the gap layer and behind the P2 pole piece using the same method as was used for the first coil layer. The third coil layer may then be formed by a conventional process. The write head is completed with the formation of the P3 pole piece to complete the yoke.